Ryan Grubb: raw computed results
The raw ranking from the production similarity calculation (face-embedding centroid cosine → gender-cohort z-normalization). The web-opinion crown, curation, and fusion re-rank are removed.
- 1
86.8%z 4.58 - 2
82.4%z 4.21 - 3
80.8%z 4.08 - 4
79.1%z 4.01 - 5
72.7%z 3.94 - 6
61.9%z 3.82 - 7
61.6%z 3.82 - 8
60.4%z 3.80 - 9
60.3%z 3.80 - 10
52.7%z 3.72 - 11
49.7%z 3.69 - 12
47.3%z 3.62 - 13
47%z 3.60 - 14
46.5%z 3.57 - 15
46.4%z 3.56 - 16
45.9%z 3.53 - 17
45.1%z 3.48 - 18
44.4%z 3.43 - 19
44.2%z 3.42 - 20
44%z 3.41 - 21
43.9%z 3.40 - 22
43.7%z 3.39 - 23
43.5%z 3.37 - 24
43.3%z 3.36 - 25
43.3%z 3.36 - 26
42.9%z 3.34 - 27
42.9%z 3.33 - 28
42.7%z 3.32 - 29
42.5%z 3.31 - 30
42.4%z 3.30 - 31
42.4%z 3.30 - 32
42.2%z 3.29 - 33
42%z 3.28 - 34
41.9%z 3.27 - 35
41.8%z 3.27 - 36
41.6%z 3.25 - 37
41.4%z 3.24 - 38
41.3%z 3.23 - 39
41.1%z 3.22 - 40
41.1%z 3.22 - 41
41.1%z 3.22 - 42
41.1%z 3.22 - 43
40.9%z 3.21 - 44
40.9%z 3.21 - 45
40.9%z 3.20 - 46
40.8%z 3.20 - 47
40.8%z 3.20 - 48
40.7%z 3.20 - 49
40.6%z 3.19 - 50
40.5%z 3.18